The valve seats in internal combustion engine heads are conventionally induction hardened surfaces which provide increased service life through reduction of wear at the valve-valve seat interface. This may be achieved by seats formed integrally within the cylinder head or by means of valve seat inserts which are retained in countersunk valve bores which open to the combustion chambers formed in the cylinder head. In particular, inserts are necessary for use with aluminum cylinder heads. Many different approaches have been taken to produce a hardened valve seat which can be inserted in a very precise accurate manner in a cylinder head on an automotive assembly line.
In one approach, the valve seat of the insert is heat treated prior to assembly and is thereafter press fitted into the valve bores of the cylinder head. The pre-hardened insert approach has generally been found to be unsatisfactory because the insert cannot be accurately positioned with the proper orientation within the valve bore. Accordingly, the insert, in a hardened state, must be machine finished or "dressed" to achieve the proper geometrical sealing relationship with the poppet valve. Such finishing operations include either surface grinding or special tooling for surface generated cutting actions. This is time consuming and costly and in turn generates heat which could, conceivably, adversely affect the hardness on the seat of the insert.
To overcome such problems, the conventional approach in use today has been to install the valve seat inserts in the cylinder heads in a metallurgically soft condition and then accurately machine the valve seat inserts by means of relatively simple machining operations to achieve the precise orientation of the insert's valve seating surface relative to the cylinder head. Thereafter, the valve seat insert is heat treated to produce the desired hardness. The heat treat operations typically use an induction heater to heat the insert to the austenitic temperature followed by rapid cooling at a rate equal to the critical cooling rate to produce a sufficient martensitic structure to meet the required hardness. Cooling at the critical rate has been conventionally achieved by using the massive cylinder head to act as a heat sink.
While this approach is fundamentally sound, in practice, and especially for those applications where cast iron inserts are inserted into aluminum cylinder heads, different rates of thermal expansion (and thermal elastic temperature limits) which exist between cast iron and aluminum could result in a partial (or even complete) loss of the insert retention force, which was produced when the insert was pressed into the countersunk valve bore, during the heating operation. The problem can be especially aggravated when it is considered that during operation of the engine, the heat, especially from the exhaust gases passing by the exhaust valve in addition to the block heat, can produce a further lessening of the retention force between the insert and the countersunk valve bore because of the different thermal expansion coefficients between the cylinder head and the insert. The problem is further aggravated when it is considered that the insert is subjected to an impact loading by the valve as it opens and closes and such shock forces could act, if the retention fit has been significantly reduced, to pry the insert loose. Accordingly, a number of approaches have been proposed to address such problems.
One such proposal, which is especially pertinent to my invention, is shown in U.S Pat. No. 4,336,432. In the '432 Patent, a significant initial retention-force is effected between the insert and the cylinder head valve bore by cryogenically cooling (usually by liquid nitrogen) the insert prior to assembly into the cylinder head bore to establish a shrink fit. Once the cryogenic has established the high preload or retention force, essentially normal utilization of induction heating apparatus at relatively short time cycles and high frequencies produces an acceptable martensitic structure on the seat of the insert while maintaining an adequate preload or retention force between the insert and the countersunk valve bore in the cylinder head. While the process shown in the '432 Patent produces a successful insert-head retention fit, it is not all together satisfactory from a commercial viewpoint, simply because cryogenic assembly is costly, time consuming and difficult to automate. The automation aspect becomes particularly troublesome when an assembly line is employed which processes both cast in place as well as separate valve seat inserts.